# EmDrive compared to flashlight in space

From my (basic) understanding the most impressive thing about the EmDrive is it violates Newton's third law. No propellant is needed, nothing is jettisoned away, yet it, allegedly, accelerates in a given direction.

I've heard that a flashlight in space, turned on, will eventually accelerate due to the energy of the light going out away from it. How different do the examples of the EmDrive and flashlight differ; is it just the EmDrive isn't even shooting light/radiation out?

In addition how could the EmDrive be much more efficient than light propulsion?

• It's unlikely that the EmDrive violates the third law, if the results are not a measurement error then it probably complies with it in ways we do not yet understand. – GdD Nov 24 '16 at 9:50

A photon drive (the flashlight example) doesn't use reaction mass, just the momentum of (massless) photons. So if it's supplied with power, e.g. from the sun, it can keep thrusting indefinitely. The most obvious example is a solar sail. However, photon drives have a thrust/power ratio of about 3.34 nN/W, which means that ~300 MW is required for 1 N of thrust.

The EmDrive is similarly free from any need for reaction mass, but it doesn't emit photons, and has a much higher thrust/power, about 1.2 µN/W with the rather crude NASA Eagleworks test article, so almost three orders of magnitude less power required for the same thrust.

An ion engine (the main competition for immediately useable high-ISP engines) has a thrust/power ratio of around 64 µN/W (Dawn's NSTARs), but has an $I_{SP}$ in the low thousands of seconds (3100 s, specifically), requiring a substantial reaction mass to go anywhere.

A VASIMR (an improved electrical propulsion drive that's not yet ready to fly) has a similar thrust/power ratio of about 70 µN/W in low gear and a similar $I_{SP}$ of a bit under 3000 s, or as little as 7 µN/W in high gear with an $I_{SP}$ of almost 30000 s.

So the EmDrive has roughly similar thrust/power as the high-gear VASIMR (intended for interplanetary travel after leaving gravity wells), but doesn't require any reaction mass, just the same power supply. So it would be cheaper to construct (without the significant fuel tanks) and quite possibly accelerate faster (with much lower initial mass). And, of course, it doesn't need to be refueled, just maintained from time to time, which makes interplanetary shuttle reuse much easier.

Of course, how this actually can be justified, physically, is the subject of the controversy. There are some wild hypotheses, which might actually pan out, but the majority of scientists are still pretty skeptical, because it mostly doesn't seem to add up.

In addition how could the EmDrive be much more efficient than light propulsion?

It probably isn't. The most likely explanation is that there is a measurement error affecting the results. The tip-off to this is that the various experimenters testing the EmDrive keep delivering different values for the thrust. It's not like one study has actually confirmed the results of another by getting the same result. We keep seeing (admittedly difficult to explain) but differing values for the efficiency of the EmDrive.

• To be fair, the most recent NASA paper makes a point of how extremely sensitive the drive's thrust is to precise RF tuning, requiring them to spend a "green" (not counted) run calibrating it before each new power setting. So the hypothesis that the different thrust values recorded are simply from different qualities of tuning is difficult to reject at this time. – Nathan Tuggy Nov 24 '16 at 0:37

Photons do have mass, albeit it is very small. They change trejectory based on the curviture of spacetime whereas neutrinos do not. The Em Drive uses radio waves which are composed of photons. But unlike the flash light where the photons are ejected as the propellant, the Em Drive the photon are directed against a plate before being allowed to escape.

• Photons do not have mass; it is neutrinos which are massive (although they are extremely light, much lighter than even an electron). – 2012rcampion Nov 24 '16 at 2:39